Text

PRA Barrier Control Program
	ML20113C508
Person / Time
Site:	San Onofre
Issue date:	06/30/1996
From:	Mcgaw J, Motamed M, Roldan Z; SOUTHERN CALIFORNIA EDISON CO.
To:	;
Shared Package
ML20113C504	List: ML20113C499; ML20113C508; ML20113C514; ML20113C519;
References
	NSG-2-3-96-001, NSG-2-3-96-1, NUDOCS 9607010286
	Download: ML20113C508 (61)
	v • d • e

.-. . _ . . .. - ---

NUCLEAR SAFETYGROUP l

PROBABILISTIC RISK ASSESSMENT BARRIER CONTROL PROGRAM San Onofre Nuclear Generating Station Units 2 is 3 JUNE 1996 NSG REPORT-2/3-96-001 i

9607010286 DR 960627 ADOCK 05000361

, PDR 1

l L NUCLEAR SAFETY GROUP 1

1 PROBABILISTIC RISK ASSESSMENT 1

BARRIER CONTROL PROGRAM f l Prepared by: M ck M. E. Motamed (NSG)

Prepared by: M b. heb Z. S. Ro,Ldan (NSG)

Reviewed By:

G Chung SG)

Reviewed By: N- - #

[/ J . W. McGaw (NEDO) j Reviewed By: c4 / u __

D. II . Calhoun (NEDO) l 4 Approved by: _R f.ise (NSG Supervisor)

/

l

i l

l l

PROBABILISTIC RISK ASSESSMENT BARR.IER CONTROL PROGRAM l

l l

I l

l

i Table of Contents

Page l

l o Acronyms 4 l o Objective 6

?

o Background 6 t

i o Scope 6 l

l o Acceptance Criterion 7 :

, o Methodology 7 l

l l

l o Results and Conclusions 8 i l l o Analysis 9 I. General Assumptions 9 l

l l II. Barrier Impairment Frequencies 11 i

l III. Barrier AOT's 12 l l

IV, Event Related Calculations 13 V. Seismically Induced Events 13 VI. Average Annual Cumulative Fuel Damage Probability 14 i

l l

t l ._. . . - . . _ . -- - . ,_ ..

Table of Contents (Cont')

o Appendix A -

Event Related Calculations and Assumptions 1 CWS Failure 16 2 Turbine Driven AFW Steam Supply Line Break 21 3 ASLB in Corridor (Unit 2) 22 4 HELB in Turbine Building 27 5 CCW/SWC Line Break in CCW area of SEB 30 6 CCW Line Break in SEB Room 017 34 7 CCW Surge Tank Failure 36 8 SDC Line Break in Mode 4 40 9 HELB Near the Roof of SEB 43 10 Fire Suppression System Line break in SEB 44 Barrier Impairment Risk Summary Table o Appendix B Piping Failure Calculations o Appendix C HRA Work Sheets i

o Appendix D Event Trees / Fault Trees and Selected Cut Sets l

i l

I l i

Acronyms l

AACFDP Average Annual Cumulative Fuel Damage Probability l AFW Auxiliary Feedwater System AOT Allowed Outage Time ASLB Auxiliary Steam Line Break l

BIP Barrier Impairment Probability BIF Barrier Impairment Frequency l CCW Component Cooling Water System i CCDP Conditional Core Damage Probability CDF Core Damage Frequency CDP Core Damage Probability CWS Circulating Water System ECCS Emergency Core Cooling System EQ Environmental Qualification ET Event Tree FT Fault Tree ,

IIRA Iluman Reliability Analysis IIELB Iligh Energy Line Break IIPSI Iligh Pressure Safety Injection IPE Individual Plant Examination LPSI Low Pressure Safety Injection i

si

I L

Acronyms (Continued)

MFLB Main Feedwater Line Break l

MSLB Main Steam Line Break NSR Non-safety Related NB Number of Barriers PCS Power Conversion System i

PRA Probabilistic Risk Assessment PSA Probabilistic Safety Assessment 1

SDC Shutdown Cooling System j SFP Spent Fuel Pool '

SG Steam Generator SR Safety Related RV Reactor Vessel RCS Reactor Coolant System SEB Safety Equipment Building j SONGS San Onofre Nuclear Generating Station SWC Salt Water Cooling System l

1 a

s

', 5 I

OBJECTIVE l

The Barrier Control Program is intended to limit to an acceptable pre-established quantitative value the increase in annual fuel damage probability attributable to barrier impairments at SONGS 2/3.

i !

i i

BACKGROUND Currently, barrier impairments are authorized on a case by case basis provided an engineering l evaluation and a successful safety evaluation (10CFR50.59) are performed prior to impairments. l This process imposes an undesirable burden on Operations, Maintenance, Engineering, and other !

organizations at SONGS. For Operations and Maintenance organizations there is an inherent !

l delay for activities requiring removal of hatches, blocking open water tight doors or breaching any

! other hazard barrier. Engineering evaluations and tracking of barrier impairments are manpower ,

intensive, requiring full time efforts of several engineers. This interim process is labor intensive, ,

costly and inefficient. i SCOPE

Amendment Applications 160 and 144 for SONGS 2 and 3, respectively, requested a new Barrier I

! Control Program Technical Specification Section 5.5.2.14, which states- 1 I

"This program provides controls for plant barriers which protect structures, systems, and l components from 1) miesiles from internal sources and adjacent buildings,2) flooding from Tsunami, internal sources and adjacent buildings, and 3) environmental hazards (such j l as steam and radiation). ., " !

l In support of the proposed Technical Specification, the long term Barrier Control Program manages barrier impairment risk associated with 1) steam release and flooding resulting from random pipe breaks, 2) steam release and flooding resulting from pipe breaks induced by seismic events up to the design basis carthquake,3) flooding resulting from tsunamis caused by seismic events up to the design basis earthquake,4) internally generated missiles, and 5) radiation hazards.

The Barrier Control Program evaluated missile and radiation hazards in the Safety Equipment !

i Building (SEB) using deterministic methods and concluded that barrier impairments had no significant effect on how these hazards impacted proximate equipment. Therefore, these hazards were not evaluated in the PRA of the SEB.

The proposed long term Barrier Control Program utilizes a combination of deterministic and probabilistic approaches. As such, the probabilistic approach is not used when a barrier l

6 !

1

impairment is supported by deterministic evaluations such as the 10CFR50.59 process, or the affected equipment is declared inoperable.

No changes are being proposed to the way in which barriers are currently being controlled to address internal fires or severe weather conditions.

No changes are being proposed to the way in which Containment barriers are currently being controlled by the Technical Specifications.

ACCEPTANCE CRITERION The Barrier Control Program is intended to limit the average increase in fuel damage probability attributable to those events listed under " Scope" resulting from barrier impairments to less than IE-6 per year per unit accrued during all modes of operation. (A permanent increase of IE-6/ year in San Onofre's fuel damage frequency would be characterized by the PSA Applications Guide (EPRI TR-105396) as "non-risk significant".) !

This acceptance criterion is an average value. While the contribution to fuel damage probability due to barrier impairments could exceed IE-6 in a given year (although unlikely due to conservatism used in the PRA models), on average this contribution will be less than IE-6/ year.

METIIODOLOGY Fuel damage likelihood attributable to barrier impairment was calculated by multiplying the frequency of those initiating events identified under " Scope" by the unavailability of one or more barriers required to ensure the operability of proximate risk-significant components and then by the conditional fuel damage probability given those components subsequently fail. This calculation required developing event trees and fault trees to model accident sequences involving j impaired barriers. Conditional fuel damage probabilities were calculated using the at-power and l shutdown PRA models of SONGS 2/3.

Initiating event frequencies are conservatively derived from generic industry data bases.

l Conditional fuel damage probabilities were calculated using the San Onofre probabilistic risk assessment models for at-power (IPE model) and shutdown conditions (Shutdown PRA model).

l Barrier unavailabilities assumed in the risk assessment were based on 1) frequencies of barrier l impairments derived from historic records, and 2) durations of barrier impairments (i.e., allowed l outage time) based on estimated future needs and risk-significance. Administrative controls will ensure average barrier unavailabilities assumed in the risk assessment remain valid by 1) ensuring adherence to specified limits on impairment durations, i.e. allowed outage times, for each barrier, 7

t l

l

and 2) providing for implementation of compensatory measures, and 3) periodic review of impairment frequencies.

The methodology was applied to the SEB to determine the average increase in the annual core damage risk associated with barrier impairments in the SEB. A scaling factor of five (5) was used to scale up the risk from SEB barrier impairments to the whole plant. The scaling factor is based on judgement and is temporarily employed until the evaluation is completed for all buildings.

Based on the SEB barrier impairment risk and the scaling factor, the average annual cumulative fuel damage risk attributable to all barrier impairments is calculated and shall meet the acceptance criterion. An iterative process was used to determine the allowed outage time (AOT) for risk significant barriers in order to meet the acceptance criterion.

The average annual cumulative fuel damage risk attributable to barrier impairments shall be l updated once the evaluation of all buildings is completed, and the AOT for risk significant barriers {

may need to be adjusted to meet the acceptance criterion.

l RESULTS and CONCLUSIONS Based on PRA, deterministic evaluations, and historical need, barrier impairment durations or allowed outage times (AOTs) were determined as follows:

EO SIGNIFICANT BARRIERS BARRIER AOT No compensatory measures implemented:

More Risk Significant Barriers 2* nours l

Less Risk Significant Barriers 12 hours1.388889e-4 days 0.00333 hours 1.984127e-5 weeks 4.566e-6 months Compensatory measures implemented:

More Risk Significant Barriers 7 days Less Risk Significant Barriers 7 days

Three barriers (removable block walls) in the Safety Equipment Building are exceptions and require compensatory measures prior to impairment.

8

l l

l i

The average increase in core damage risk attributable to barrier impairments associated with the Safety Equipment Building have been calculated to be 6.2E-8 per year. A factor of five (5) was temporarily used to scale up to the fuel damage risk for the whole plant. The average annual cumulative fuel damage risk attributable to all barrier impairments in the plant was estimated to be 3.lE-7, which meets the acceptance criterion of IE-6/ year.

ANALYSIS Event trees were developed for three cases :

o Modes 1 to 4 with SG available, referred to as Modes 1-4.

o Shutdown Cooling System operation (2 trains of SDC operable) in Modes 4,5, and Mode 6 with RCS water level less than 23 ft above the RV flange , referred to as " Modes 4,5,6 (level <23')" (See Assumption 3 regarding leve1<l2' below.)

o Shutdown Cooling System operation in Mode 6 with water level at 23 ft above RV flange or greater (one train of SDC Operable), referred to as " Mode 6 (level >23')". (See Assumption 3 regarding level <l2' below.)

For the above three cases, system configurations were selected to bound all modes of operations.

For example, in Modes 1 to 4, Main Feedwater pumps were not credited to mitigate events, because the pumps may not be available in Mode 4. In Modes 4 to 6, mid-loop operation with high decay heat was used, because it is the highest risk plant configuration in these modes.

I. General Assumptions:

l l

1. Based onjudgement, a factor of 5 was used to scale up the risk of core damage l attributable to barrier impairments in the SEB to the whole plant. This is a temporary l l means to ensure that the cumulative risk from impairments in the plant meets the acceptance criterion of IE-6/yr. The scaling factor will not be used in the final calculation of fuel damage risk when all the buildings are analyzed.

2. Engineering judgement was used to exclude steam release paths which could be considered ineffective in propagating steam to target areas creating a significant change in ambient environmental parameters. These are called "tonuous paths" and are identified in the assumptions for each event.

3. During SDC operation when refueling cavity water level is less than 23 ft above the RV flange, two trains of SDC are assumed to be operable. When the cavity level is 23 ft above the flange or higher, one train of SDC is assumed to be operable. The 23 ft level will be changed to 12 ft when a proposed Tech. Spec. change (Proposed Change Number 9

PCN-458) is approved by the NRC. To be conservative, the time to core damage for the water level initially at 12 ft above the RV flange was used in the event trees.

4. During Modes 4, 5, and 6 with water level less than 23 feet above the RV flange, consistent with the outage planning, three charging pumps are available for injection to the ,

RCS. This assumption will be modified to water level less than I' feet above the RV ,

flange when a proposed Tech. Spec. change (PCN-458)is approm.

5. During Modes 4, 5, and Mode 6 with RCS level <23' above RV flange, on the average, l

two of three charging pumps are adequate to make up for inventory loss due to RCS boil offin the event ofloss of SDC. This assumption will be modified to water levelless than 12 feet above the RV flange when a proposed Tech. Spec. change (PCN-458)is approved.

6. Pipe failure probabilities provided in EPRI TR-102266, " Pipe Failure Study Update", were used for safety related (SR) pipes. Failure probability of non-safety related (NSR) piping were assumed to be ten times higher than SR pipes of the same category as defined in the EPRI report.

7. Although the Swing CCW/HPSI pumps transfer switches are expected to survive steam environment (<160 F), it was conservatively assumed that steam release in the common areas of SEB would cause the swing CCW or HPSI pumps to fail due to transfer switch failure.

8. Except where noted otherwise, steam release to the ECCS common area of SEB (e.g.,

room 017) would bypass Train A ECCS pump room barriers through the shutdown cooling tunnel resulting in failure of Train A ECCS pumps.

9. Steam release to the SEB room 017 does not fail CCW or SDC Systems for 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months .

This was based on engineering evaluation. (

References:

EQDP M37703, M37618; and Memo. for file by M. Plouff and R. Wise, dated 12/1/95) 1

10. Removable block walls in the CCW area of the SEB were treated as one barrier because l their combined impairment frequencies were less than other barriers in the locked / secured category.

11. Steam release to CCW surge tank rooms does not fail the CCW system for 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months . This was based on engineering evaluation. (Reference EQDP M37606, E-mail Memo. from M.

Plouff to M. Motamed dated 5/21/96) l i

10

4 12 Steam release to the SDC heat exchanger rooms does not fail SDC system for 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months . .

SDC valves in the area would not fail in steam environment, and failure of flow transmitters in the area does not fail SDC operation for 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months . (The SDC valves evaluation is documented in EQDP M37708)

13. Steam release to the CCW heat exchanger and piping area does not fail the CCW and SWC systems for 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months . This was based on engineering evaluation. (

References:

EQDP M37703, M37618; and Memo. for file by M. Plouff and R. Wise, dated 12/1/95)

14. This evaluation does not include possible impact of hazard propagation through HVAC systems pending the outcome of on-going deterministic evaluations.

i 15. Missile hazards were not included in the barrier PRA pending outcome of on-going deterministic evaluations.

16. Penetration seal impairments were not included in the PRA and will be evaluated separately. However, the scope of the PRA will be expanded to include risk significant penetration seals.

17. ~ Barrier impairment frequencies were estimated using impairment data for a selected sample of safety significant barriers in the SEB compiled by the SONGS Emergency Preparedness Fire Protection Group from 1988 to 1995.

18. The safety equipment in the SEB elevations 30', 50', and 70' were qualified for steam environment based on an engineering evaluation (

References:

EQ Master List, DBD-S023-TR-EQ).

19. Control of simultaneous (multiple) barrier impairments will be in place to ensure the applicability of this risk analysis.

20. Drain lines were considered to be tortuous steam release paths and are therefore excluded.

21. Tsunami barrier impairments will be included in this evaluation prior to implementation of the proposed barrier control program.

1 II. Barrier impairment Freauencies !

h Barriers were divided into two groups: locked / secured barriers (e.g., secured hatches, locked doors, and removable block walls), and unlocked / unsecured barriers (e.g., unlocked doors, card key operated doors). Impairment frequency data were further broken down for power operation i

i l1 l

t 4

o 4

J (defined for this purpose as Modes 1-4 with steam generators functional), and outages (SDC j operations, Modes 4-6) creating four impairment categories:

1. Unlocked / unsecured barriers - power operation.

) 2. Locked / secured barriers - power operation.

3. Unlocked / unsecured barriers - outages. I l

j 4. Locked / secured barriers - outages.

{

1

Barrier impairment frequencies were estimated using SONGS 2/3 data for a selected sample of l

SEB barriers compiled by the Emergency Preparedness Fire Protection Group from 1988 - 1995.

Average impairment frequencies for barriers in the above 4 categories were estimated using actual

)

, impairment data for 15 locked barriers and 12 unlocked barriers associated with the Safety l Equipment Building. Impairment frequencies averaged for all 27 barriers was also calculated to '

apply to cases where a combination of secured and unsecured barriers were evaluated. The results are :

.i

{ Locked / Secured Unlocked / Unsecured Average Impairments /yr j Impairments /yr Impairments /yr Power Operations 0.4 2.5 1.3 !

Outage 0.6 7.2 3.5 1 ,

I III. Barrier AOT l

1 The risk significance of barriers in the SEB was determined by a sensitivity study. Bamers protecting against CWS flooding were identified as risk significant and assigned an AOT of 2

, hours. The 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months was the result of an iterative process to minimize the contribution of risk significant barriers to the overall barrier impairment risk. Compensatory measures were identified I which could extend the two hour AOT. Less risk significant barriers protecting safety equipment in the SEB against hazards were assigned AOTs of 12 hours1.388889e-4 days 0.00333 hours 1.984127e-5 weeks 4.566e-6 months based on historic need even though ,

PRA could support longer AOTs. In some cases compensatory measures have been identified to '

extend the 2 and 12 hours1.388889e-4 days 0.00333 hours 1.984127e-5 weeks 4.566e-6 months AOTs for hazard barriers to 7 days. This risk analysis did not credit 4

compensatory measures for less risk significant barriers.

l

\

I i

l 4

i 12

IV. Event Related Calculations The following flooding and steam hazards and barrier impairment related events were evaluated for the SEB using PRA:

1. Circulating Water System failure.

2. Turbine Driven Auxiliary Feedwater Pump turbine steam supply line break in the Aux. Feedwater pump room.

3. Auxiliary Steam line break in the corridor south of Unit 2 SEB.

4. High energy line break (HELB) in the Turbine Building below turbine deck.

5. Component Cooling Water / Salt Water Cooling systems piping failure in the CCW area of SEB.

6. CCW piping failure in SEB room 017.

l

7. CCW System Surge Tank rupture.

8. Shutdown Cooling System (SDC) piping failure in Mode 4, RCS temperature

>212 F.

9. HELB Near the roof of the SEB. I

10. Fire Suppression System piping failure in SEB.

Other events were evaluated deterministically. Calculations to support the event trees are provided in Appendix A, pipe failure calculations (initiator frequencies) are provided in Appendix B, and Human Reliability Analysis (HRA) work sheets for five operator actions from a total of eleven work sheets are provided in Appendix C. Appendix D provides all event trees and fault trees developed for the SEB barrier impairment risk analysis, and cutsets for a selected number of 1 event trees.

V. Seismically Induced Events The impact of seismic events up to the design basis earthquake was evaluated. The events which could impact safety equipment in the SEB were determined to be seismic induced HELB in the turbine building or SEB corridor, seismic induced CWS failure, and seismic induced Tsunami.

Based on seismic evaluations, the high energy lines within range of the SEB were determined to 13

( have a negligible probability of failure due to a design basis seismic event. (Average Spectral Acce'eration = 3.15g, Beta, = 0.3, Beta , = 0.45) (Reference, Repon NSG 96-002, dated 2/96).

The CWS failure due to a seismic event could cause flooding in the turbine bt ilding. However, if off site power is concurrently lost, the circulating water pumps will lose power and a significant flood can not occur. A special study was performed that demonstrated the probability of a seismic induced CWS failure with offsite power available was negligible. (

Reference:

SCE letter to the NRC," Response to GL 88-20, IPEEE". dated December 15,1995.)

As stated earlier, Tsunami barrier impairments will be included in this evaluation prior to implementation of the proposed barrier control program.

l VI. Average Annual Cumulative Fuel Damage Probability The cumulative core damage risk due to barrier impairments are calculated by summing the CDF l in the event trees for each of the three plant configurations:

Sum of Fuel Damage l Prob. From Event Trees I o Modes 1-4: XI = 1.58E-10/ day i

o Modes 4, 5, 6(level <23'): X2 = 2.17E-10/ day l

I o Mode 6(level >23'): X3 = 3.42E-10/ day For a typical 18 month fuel cycle at SONGS 2/3 followed by a 60 day refueling outage, the fraction of time in which the plant operates in the above configurations is approximately:

Fraction of l Operating time o Modes 1-4 Y1 = 0.9 l

l o Modes 4, 5, 6(Ievel<23') Y2 = 0.0 i o Mode 6(level >23') Y3 = 0.05 l

14

I l

l l'

The annual average cumulative fuel damage probability due to barrier impairments in the SEB is calculated by multiplying the cumulative fuel damage frequency in each plant configuration by the l fraction of time the plant operates in that configuration and sum over all configurations.

1 AACFDP (SEB) = (X1 x Y1 + X2 x Y2 + X3 x Y3) x 365 = 6.2E-8 l

- A factor of 5 is used to scale up the SEB to the whole plant:

AACFDP (Plant) = AACFDP(SEB) x 5 = 3.1E-7 The acceptance criteria of IE-6/yr is therefore satisfied. The scaling factor is a temporary estimate and will not be used in the final calculation of the fuel damage risk. As stated earlier, the ,

AACFDP will be calculated for all areas of the plant prior to implementing the long term barrier I control program. If a 24 month fuel cycle is used, the AACFDP would be slightly lower than what is calculated above.

l l

l I

i

15

l l

APPENDIX A i

i l

1 i

l EVENT RELATED CALCULATIONS AND ASSUMPTIONS i

1 l

l l

I l

l l

l 1

6 l

EVENT RELATED CALCULATIONS AND ASSUMPTIONS Pipe failure calculations (initiator frequencies) used in this Appendix are provided in Appendix B, and Human Reliability Analysis (HRA) work sheets for five operator actions from a total of eleven work sheets are provided in Appendix C. Appendix D provides all event trees and fault trees developed for the SEB barrier impairment risk analysis, and cutsets for a selected number of event trees.

1. Circulating Water System Failure l Circulating Water System (CWS) failure resulting in flooding in the turbine building is l evaluated. CWS flooding in one unit will impact both units. If certain barriers were l impaired, the flooding may enter the SEB and fail equipment. The flooding could cause a l turbine trip.

I

! SEB primary barriers associated with the Circulating Water System flooding were determined to be risk significant and assigned AOTs of 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months without flood watch and 7 j days with flood watch. These AOT's apply to the flood hazard.

1 Assumptions:

1.1 Turbine Building flooding due to Circulating Water System expansionjoint failure at the condenser has a frequency of 6.2E-3/yr. (

Reference:

" Internal Flood l Frequencies during Shutdown and Operation for Nuclear Plants, " N. O. Siu, et.

al., prepared for Public Service of New Hampshire, PLG, Inc., PLG-0624, May j 1988.)

l 1.2 Based on transient flood calculations, in the event of a Circulating Water System flood in the turbine building with barriers to the SEB impaired, if the Circulating l Water pumps are tripped in 10 minutes, the transient flood level does not result in j the failure of safety equipment in the SEB. The SEB block walls are assumed not to be voluntarily impaired below elevation 9'0" (unless a berm is first installed up to elevation 9'0").

l The following ET's were developed:

CWS flood in Modes 1-4 with and without flood watch (Figures 1.a and 1.b).

CWS flood in Modes 4, 5, 6(Level <23') with /without flood watch (Figures 1.c, l .d).

W 16

I CWS flood in Mode 6(level >23') with/without flood watch (Figures 1.e,1.1).

Calculations for Figure 1.a:

SEB flooding due to CWS failure in Modes 1-4, barriers impaired without flood watch.

Initiating event frequency, Node IE = 3.4E-5/ day Where, IE = CWS failure rate for both units per day (6.2E-3 per yr/365) x 2 = 3.4E-5/ day Node OP = SE-1 (see HRA Work Sheet 1)

Node DR = BIFxNBxAOT/(365 x 24) = 2.7E-4 Where, DR = Probability of SEB barriers 008,009, or ECCS area block wallimpaired BIF = 0.4 Barrier impairment frequency for secured barrier, l Modes 1-4 NB = 3 Number of barriers i AOT = 2 hrs Barrier allowed outage time (AOT) Without flood l

watch !

Node DRA = BIFxNBxAOT/(365 x 24) = 2.7E-4 Where, DRA= Prob. Of SEB barriers 007, CCW area block wall, or berm in the corridor impaired BIF = 0.4 Secured barrier, Modes 1-4 NB = 3 AOT = 2 hrs Without flood watch Node CDA = PRA computer run, CCDP given a loss of ECCS and a CWS flood.

Node CDB = PRA Run, CCDP given a loss of CCW and a CWS flood. j I

Calculation for Figure 11.,

SEB flooding due to CWh failure in Modes 1-4, barriers impaired with flood watch..

Initiating event frequency, Node IE = 3.4E-5/ day Where, IE = CWS failure rate for both units per day 17

I I

l (6.2E-3 per yr/365) x 2 = 3.4E-5/ day Node OP = SE-2 (see HRA Work Sheet 2)

Node DR = BIFxNBxAOT/365 = 2.3E-2 Where, DR = Probability of SEB barriers 008,009, or ECCS area block

wallimpaired BIF = 0.4 Secured barrier, Modes 1-4 NB = 3 AOT = 7 days With flood watch Node DRA = BIFxNBxAOT/365 = 2.3E-2 Where, DRA= Prob. Of SEB barriers 007, CCW area block wall, or berm in the corridor impaired BIF = 0.4 Secured barrier, Modes 1-4 NB = 3 AOT = 7 days With flood watch !

l Node CDA = PRA Run, CCDP given a loss of ECCS and a CWS flood.

Node CDB = PRA Run, CCDP given a loss of CCW and a CWS flood.

Calculations for Figure 1.c SEB flooding due to CWS failure in Modes 4, 5, and 6(Level <23'), barriers impaired with flood watch.

Initiating event frequency, Node IE = 2.4E-5/ day Where, IE = CWS failure rate for both units per day (6.2E-3 per yr/365) x MUL = 2.4E-5/ day MUL = 1.4 Operating unit's CWS plus the shutdown units CWS only available 40% of the time Node SEB = BIFxNBxAOT/365 = 7E-2 Where, Node SEB = Probability of SEB barriers 007,008,009, SEB block walls, or berm in the corridor impaired BIF = 0.6 Secured barriers, Modes 4, 5,6 NB = 6 3 doors,2 block walls, and one berm.

l 18

1 l l

\

l AOT = 7 days With flood watch l l

f Node OP = SE-2 (see HRA Work Sheet 2) l -

I I Node OPl = IE-3 (see HRA Work Sheet 3) l l Node CHG = PRA Run,2 of 3 charging pumps fail to operate for 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months . ,

l I Calculations for Figure 1.d l

l

]

SEB flooding due to CWS failure in Mode 4, 5, and 6(Level <23'), barriers impaired without flood watch.

Initiating event frequency, Node IE = 2.4E-5/ day Where, IE = CWS failure rate for both units per day l (6.2E-3 per yr/365) x MUL = 2.4E-5/ day l MUL = 1.4 Operating units CWS plus the shutdown units CWS only available 40% on the time ,

l Node SEB = BIFxNBxAOT/365 = 8.2E-4 l l

Where, Node SEB = Probability of SEB barriers 007, 008, 009, SEB l block walls, or berm in the corridor impaired l BIF = 0.6 Secured barriers, Modes 4,5,6 l NB = 6 3 doors,2 block walls, and one berm. 1 1

AOT = 2 hrs Without flood watch Node OP = SE-1 (see HRA Work Sheet 1)

Node OPl = IE-3 (see HRA Work Sheet 3)

Node CHG = PRA Run,2 of 3 charging pumps fail to operate for 24 hrs.

Calculations for Figure 1.e l SEB flooding due to CWS failure in Mode 6(Level >23'), barriers impaired with flood I

watch. .

Initiating event frequency, Node IE = 2.4E-5/ day i Where, IE = CWS failure rate for both units per day l (6.2E-3 per yr/365) x MUL = 2.4E-5/ day i

19 f

I

l MUL = 1.4 Operating unit's CWS plus the shutdown units CWS only available 40% of the time Node SEB = BIFxNBxAOT/365 = 7E-2 ,

l Where, Node SEB = Probability of SEB barriers 007,008,009, SEB l block walls, or berm in the corridor impaired i BIF = 0.6 Secured barriers, Mode 4,5,6 !

NB = 6 3 doors,2 block walls, and one berm. !

AOT = 7 days With flood watch l

Node OP = SE-2 (see HRA work sheet 2) i Node OPl = 1.5E-3 (see HRA work sheet 4)

Node SFP = 8E-4 l

Where, SFP = Mechanical failure to open/close eight critical valves to align l SFP to cool RCS and to cross tie CCW to the other unit. l l

In the event of a loss of SDC, the operator will align Spent I Fuel Pool Cooling to recirculate the RCS, and if CCW failed, align the SFP heat exchangers from the opposite Unit CCW. To place the SFP pumps in the RCS recirculate mode, four critical valve alignments are required: 1) open 1201MU033, 2) open 1219MU013, 3) close 1219MU012, and 4) open 120lMU994. If CCW in the affected unit also failed, to cross tie to the other unit's CCW system, four more critical valve alignments would be required: 5) open HV-6217, 6) open HV-6465, 7) open SA 1203MU002, and

8) open SA 1203MU100. Operator actions and procedures are discussed in HRA work sheet 4 in Appendix C.

Mechanical failure of any of these eight valves was assumed to result in failure of the required action. Manual valve failure to open or close (i.e.,lE-4/ demand, NUREG/CR-4550) was conservatively applied to all six valves (i.e.,

8xIE-4 = 8E-4/d).

Calculations for Figure 1.f SEB flooding due to CWS failure in Mode 6(Level >23'), barriers impaired without flood 20

watch.

Initiating event frequency, Node IE = 2.4E-5/ day Where, IE = CWS failure rate for both units per day l (6.2E-3 per yr/365) x MUL = 2.4E-5/ day !

MUL = 1.4 Operating units CWS plus the shutdown units CWS only available 40% on the time Node SEB = BIFxNBxAOT/365 = 8.2E-4 l

Where, Node SEB = Probability of SEB barriers 007,008,009, SEB l block walls, or berm in the corridor impaired i BIF = 0.6 Secured barriers, Modes 4, 5,6 l NB = 6 3 doors,2 block walls, and one berm.

AOT = 2 hrs Without flood watch Node OP = SE-1 (see HRA Work Sheet 1)

Node OPl = 1.5E-3 (see HRA Work Sheet 4)

Node SFP = 8E-4 ;

Where, SFP = Mechanical failure to open/close eight critical valves to align SFP to cool RCS and to cross tie CCW to the other unit.

See discussion for Figure 1.e above.

2. Turbine Driven Auxiliary Feedwater Steam Supply Line Break A break in the Steam supply piping to the Aux. Feedwater (AFW) pump turbine in the AFW pump room could release steam to the AFW piping tunnel. The steam could enter the SEB if barriers 010 or 011 were impaired. This event only applies to Modes 1-4 where steam generators are available.

Assumptions:

2.1 In the event of a steam lir.e break in the AuxHiary Feedwater (AFW) Pump room, if the break is isolated within 30 minutes, the motor driven pumps are assumed to remain operable. If the break is not isolated in 30 minutes all AFW pumps are assumed to fail. This is based on engineering analysis. (Reference, SCE Design Calculation DC#N-4090-009).

21 l

l

Calculations for Figure 2.a:

An ET was developed for the AFW turbine steam supply piping failure in the AFW pump room, Modes 1-4 (see Figure 2.a). The event tree was simplified by not crediting SEB door 012 as a barrier, and redundant CCW/SWC trains.

Initiating event frequency, Node AFW = 2.14E-7/ day Where, AFW = Pipe failure frequency

= # sections x pipe failure rate per section Node BRR = BIFxNBxAOT/(365 x 24) = 1E-3 Where, BRR = Prob. ofimpairment of SEB barriers 010, or 011

BIF = 0.4 locked door, Modes 1-4 NB = 2 two locked doors AOT = 12 hrs for steam barriers to common areas Node SLB = IE-2 (see HRA work Sheet 5)

Node DRS = (BiFi xNBi xAOT/365) x (BIF2xNB 2xAOT/365) = 9.6E-3 Where, DRS = Prob. ofimpairment of SEB doors 015 or 016, and the CCW trains A and B pump room barriers.

BIF i= 2.5 unlocked doors, Modes 1-4 AOT = 7 days internal barriers NB, = 2 doors 015,016 BIF 2= 1.3 Average BIF for CCW pump room door and hatch NB 2= 4 one door and one hatch in each of two pump rooms Node CDA = PRA Run, CCDP given loss of ECCS and TDAFW pump.

Node CDB = PRA Run, CCDP given loss of CCW and TDAFW pump.

Node CDC = PRA Run, CCDP given loss of CCW and AFW Systems.

Node CDD = PRA Run, CCDP given loss ofECCS and AFW system.

i

3. Auxiliary Steam Line Break (ASLB)in Corridor (Unit 2 only)

This event is a postulated break in the 10" Auxiliary Steam line in the corridor south of Unit 2 SEB. This event applies to all Modes of operation except when the line is isolated.

I 22 i

i Event trees were simplified by not crediting Swing CCW pump, and assuming failure of one train of CCW causes failure of the other train of CCW.

The following event trees were developed to evaluate all plant conhgurations:

ASLB in Modes 1-4, Figure 3.a.

ASLB in Modes 4, 5, 6(level <23'), Figure 3.b.

ASLB in Mode 6(level >23'), Figure 3.c.

Assumptions:

3.1 ~ Auxiliary SLB in the corridorjust south of Unit 2 SEB can enter the SEB only through doors 017,018, the block walls between CCW area and the turbine building, and door C2-103 and SDC tunnel. Steam release through C2-102 and the SDC tunnel to the ECCS area of SEB wasjudged to be tortuous path and therefore not considered.

Calculations for Figure 3.a:

ASLB in the corridor south of Unit 2 SEB, in Mode 1-4. To simplify this event tree, it was conservatively assumed that barrier impairments (door or hatch) allowing steam to enter either train A or B CCW pump rooms would result in a total loss of CCW.

Initiating Event, Node SL1 = 2.14E-7/ day Where, SLI = # pipe section x failure rate per section Node DR = BI(exterior) x BI(interior) = 7.4E-4 Where, DR = Prob. of SEB barriers 017, 018, or block wall in the CCW area are impaired; and CCW trains A/B pump room barriers are breached.

BI(exterior) = (BIFi xAOT + BIF 2xNB 2xAOT )/(365 2 x 24)

= (0.4 x 12 + 2.5 x 2 x 12)/(365 x 24) = 7.4E-3 BI(interior)=(BIF3x NB 3x AOT 3/365)

= ((1.3 x 4 x 7)/365) = 0.1 BI(exterior) = Prob. of SEB barriers 017,018, or CCW area block wall are impaired.

23

BIF, = 0.4 CCW area block wall (secured barrier), Mode 1-4 AOT i= 12 hours1.388889e-4 days 0.00333 hours 1.984127e-5 weeks 4.566e-6 months steam barrier to common area BIF 2= 2.5 unlocked doors 017,018, mode 1-4 AOT 2= 12 hours1.388889e-4 days 0.00333 hours 1.984127e-5 weeks 4.566e-6 months steam barrier to common area NB2= 2 two doors BI(interior) = Pro. ofimpairment of CCW trains A and B pump room barriers.

BIF 3= 1.3 average BIF for a CCW pump roc:n door / hatch NB 3= 4 two doors and two hatches AOT 3= 7 days internal SEB barriers Node DRA = BIFxAOT/(365 x 24) = 3.4E-3 Where, DRA = prob. of door C2103 impaired BIF = 2.5 unlocked door C2103, Modes 1-4 AOT = 12 hrs steam barrier to common area Node DRB = BI, x BI2= 1.5E-3 Where, DRB = Prob. of hatches to the SDC tunnel impaired, and CCW trains A/B pump room barriers are breached.

BI, = BIFi xNBi xAOT /365

= (0.4 x 2 x 7)/365 = 1.5E-2 BI2= (BIF 2x NB 2x AOT 2/365)

= ((1.3 x 4 x 7)/365) = 0.1 BI, = Prob. ofhatches impaired to SDC tunnel.

BIF, = 0.4 two SDC tunnel hatches, Mode 1-4 AOT = 7 days i

BI 2= Pro. ofimpairment of CCW trains A and B pump room barriers.

BIF 2= 1.3 average BIF for a CCW pump room door / hatch NB 2= 4 two doors and two hatches AOT 2= 7 days internal SEB barriers l

Node CD = PRA Run, CCDP given a loss of ECCS and CCW Swing pump.

Node CDX = PRA Run, CCDP given a loss of CCW.

Calculations for Figure 3.b:

24

ASLB in the corridor south of Unit 2 SEB, Modes 4, 5,6(level <23'). To simplify this event tree, it was conservatively assumed that barrier impairments allowing steam to enter either train A or B CCW pump rooms would result in a totalloss of CCW.

Initiating Event, Node IE = 2.14E-7/ day Where, IE = # pipe section x failure rate per section Node DR = DI(exterior) x BI(interior) = 5.5E-3 Where, DR = Prob. of SEB barriers 017, 018, or block wall in the CCW area are impaired; and CCW trains A/B pump room barriers are breached.

BI(exterior) = (BIFi xAOT + BIF 2xNB 2xAOT 2)/(365 x 24) i

= (0.6 x 12 + 7.2 x 2 x 12)/(365 x 24) = 2.04E-2 Bi(interior) = (BIF3 x NB 3x AOT 3/365)

= ((3.5 x 4 x 7)/365) = 0.27 BI(exterior) = Prob. of SEB barriers 017,018, or CCW area block wall are impaired.

BIF, = 0.6 CCW area block wall (secured barrier), Mode 4,5,6(level <23')

AOT i= 12 hours1.388889e-4 days 0.00333 hours 1.984127e-5 weeks 4.566e-6 months steam barrier to common area BIF2 = 7.2 unlocked doors 017,018, Mode 4,5,6(level <23')

NB 2= 2 two doors BI(interior) = Pro. ofimpairment of CCW trains A and B pump room barriers.

BIF3= 3.5 average BIF for a CCW pump room door / hatch, Modes I 4,5,6(level <23')

NB 3= 4 two doors and two hatches AOT 3= 7 days internal SEB barriers 1 Node DRA =BIFxAOT/(365 x 24) = IE-2 l

Where, DRA = prob. of doors C2103 impaired l BIF = 7.2 unlocked door C2103, Modes 1-4 '

l AOT = 12 his steam barrier to common area

! Node DRB = Bi + BI: = 1.5E-1 1

Where, DRB= Prob. of door SE001 impaired, or SDC tunnel hatches open and l CCW Train A/B pump room barriers impaired {

l l

l 25 l 1

BI i= BIF ixAOT/365 =1.4E-1 BI 2= (BIF 2xNB 2xAOT/365)x(BIF 3xNB 3xAOT/365) = 6.2E-3 BI, = prob. of door SE001 impaired BIFi = 7.2 door SE001 AOT = 7 days BI 2= Prob. of SDC tunnel hatches and CCW pump room barriers ;

impaired l BIF2= 0.6 SDC hatches NB2 = 2 two hatches !

BlF3 = 3.5 CCW pump room doors and hatches NB3 = 4 two doors and two hatches Node ALI = IE-3 (see HRA work Sheet 3)

Node CHR = PRA Run,2 of 3 charging pumps fail to operate for 24 hrs.

Calculations for Figure 3.c: !

l ASLB in the corridor south of Unit 2 SEB, Mode 6(level >23'). To simplify this event tree, it was conservatively assumed that barrier impairments allowing steam to enter either l train A or B CCW pump rooms would result in a total loss of CCW.

The bounding case is when Train A SDC is operable. In this case, impairing door C2103 )

would result in the failure of the SDC System, because Train A ECCS pump room directly communicates with the SDC tunnel..

Initiating Event, Node IE = 2.14E-7/ day Where, IE = # pipe section x failure rate per section Node DR = BI(exterior) x BI(interior) = 5.5E-3 Where, DR = Prob. of SEB barriers 017,018, or block wall in the CCW area are ,

impaired; and CCW trains A/B pump room barriers are breached. l BI(exterior) = (BIF i xAOT, + BIF 2xNB 2xAOT 2)/(365 x 24) j

= (0.6 x 12 + 7.2 x 2 x 12)/(365 x 24) = 2.04E-2 ;

i BI(interior)=(BIF 3x NB 3x AOT 3/365) ;

= ((3.5 x 4 x 7)/365) = 0.27 BI(exterior) = Prob. of SEB barriers 017,018, or CCW area block wall are 26

~

l impaired.

BIF, = 0.6 CCW area block wall (secured barrier), Mode 6 l AOT , = 12 hours1.388889e-4 days 0.00333 hours 1.984127e-5 weeks 4.566e-6 months steam barrier to common area BIF2 = 7.2 unlocked doors 017,018, Mode 6 ,

l AOT 2= 12 hours1.388889e-4 days 0.00333 hours 1.984127e-5 weeks 4.566e-6 months steam barrier to common area !

l NB2= 2 two doors l BI(interior) = Pro. ofimpairment of CCW trains A and B pump room l barriers.

l BIF3= 3.5 average BIF for a CCW pump room door / hatch, Modes 6 l NB 3= 4 two doors and two hatches AOT 3= 7 days internal SEB barriers 1

Node DRA = BIFxAOT/(365 x 24) = 1.0E-2 I I

Where, DRA = prob. Of door C2103 impaired l

BIF = 7.2 unlocked door C2103, Modes 1-4 AOT = 12 hrs steam barrier to common area l 4

l Node OP = 1E-3 (see HRA work Sheet 4)

Node SFP = 8E-4 1

Where, SFP = Mechanical failure to open/close eight critical valves to align ;

SFP to cool RCS and to cross tie CCW to the other unit.

See discussion for Figure 1.e above.

I 1

4. High Energy Line Break in Turbine Building High Energy Line Break (HELB) in the turbine building including Main Steam, Main Feed, Steam Generator Blowdown, and Auxiliary Steam piping failure is evaluated.

Assumotions:

4.1 In the event of a HELB in the turbine building, steam release through the doors C2-103 and C2-102 through the SDC tunnel to the SEB wasjudged to be tortuous path and therefore not considered.

4.2 A HELB in Unit 2 Turbine Building does not impact the SEB of Unit 3 and sice versa.

4.3 During Shutdown Cooling System operation in Modes 4,5, and 6, the only high 27

energy piping in the Turbine Building that could impact the SEB is Auxiliary Steam piping.

4.4 HELBs west of the condenser and turbine housing does not impact SEB due to buoyancy effects and distance.

The following event trees were developed:

l -

HELB in the Turbine Building, Modes 1-4, Figure 4.a.

Auxiliary Steam Line Break (ASLB) in the Turbine Building, Mode 4, C. and Mode 6(level <23'), Figure 4.b.

ASLB in the Turbine Building, Mode 6(level >23'), Figure 4.c.

! Calculation for Figure 4.a:

HELB in the Turbine Building Modes 1-4. Main Feedwater System was not credited for recovery since it may not be available in Mode 4.

Initiator frequency, Node IE = 1.9E-5/ day l

Where, IE = HELB frequency (# piping sections x Pipe failure rate per section)

(MFLB, MSLB, ASLB, and Blowdown LB) I l

Node A = Probability of barrier impairments resulting in steam release due to HELB to reach the CCW A pump room (see FT attached to Fig. 4.a).

Node B = Probability of barrier impairments resulting in steam release due to HELB to reach the CCW B pump room (see FT attached to Fig. 4.a).

Node C = Probability of barrier impairments resulting in steam release due to HELB ,

to reach the ECCS B pump room (see FT attached to Fig. 4.a).

Node D = Probability of barrier impairments resulting in steam release due to HELB to reach the ECCS A pump room (see FT attached to Fig. 4.a).

l- Node E = PRA Run, CCDP given a loss of ECCS A and loss of main feed. l Node F = PRA Run, CCDP given a loss of ECCS and loss of main feed.

Node G = PRA Run, CCDP given a loss of CCW B and loss of main feed.

I 28

1 i

Node H = PRA Run, CCDP given loss of CCW B, ECCS A and main feed.

Node I = PRA Run, CCDP given loss of CCW A and main feed.

I Node J = PRA Run, CCDP given loss of CCW A, ECCS B, and main feed. ;

Node K= PRA Run, CCDP given loss of CCW and main feed.

l Calculations for Figure 4.b:

HELB in Modes 4,5,6(level <23'). In these Modes, the only HELB is the ASLB. j Initiating event, Node IE = 3.2E-6/ day Where, IE = Frequency of Auxiliary Steam piping failure in the Turbine Building. (# piping sections x Pipe failure rate per section)

Node A = Prob. of barrier impairments resulting in steam release to ECCS pump )

rooms or CCW B pump room failing ECCS A and ECCS B, or CCW (see FT attached to Figure 4.b).

l Node B = Prob. Of barrier impairments resulting in steam release to ECCS A pump room or CCW A pump room (see FT attached to Fig. 4.b).

Node C = SE-2 (see HRA work sheet 10) i Node D = IE-3 (see HRA work sheet 3)

Node E = PRA Run,2 of 3 charging pumps fail to operate for 24 hrs. l i

Calculations for Figure 4.c:

l The only HELB in Mode 6(level >23') is the Aux. Steam Line Break. In this Mode one l train of SDC is removed from service for mamtenance. Worst case of barrier impairments '

is when Train B ECCS is out of service, because steam release to ECCS common area can bypass ECCS A pump room barriers.

Initiating event, Node IE = 3.2E-6/ day Where, IE = Frequency of Auxiliary Steam piping failure in the Turbine Building. (# piping sections x Pipe failure rate per section) 29

l Node STM = Prob. Of barriers impairments resulting in steam release to ECCS A pump room or CCW A pump room (see page 3 of FT attached to Figure 4.b).

Node OP = 1.5E-3 (see HRA work sheet 4)

Node SFP = 8E-4 Where, SFP = Mechanical failure to open/close eight critical valves to align SFP to cool RCS and to cross tie CCW to the other unit.

See Section 1.c above.

5. CCW/SWC Line Break in CCW Area of SEB Line breaks in the CCW and SWC piping in SEB rooms 009 to 014 and 022 to 026 are evaluated. The barriers associated with the failed train of CCW/SWC are ignored.

There are drain lines from CCW pump rooms to a sump in the CCW common area of the SEB. In the event of flooding in the common area, the check valves in the drain lines are not credited, in the design basis, to prevent flooding the CCW pumps. Therefore, the water tight doors to the pump rooms are not considered as flood barriers in the design I

basis of SONGS 2/3. However, this analysis conservatively assumes these doors to be l

flood barriers, as stated in assumption 5.3 below.

l Assumotions:

5.1 The design basis flooding calcolations for the CCW and SWC line breaks in the SEB assumed the break is isolated in 30 minutes. l l

5.2 Whenever the hatch over a CCW pump room is opened, the associated flood door ;

is opened. This will prevent flooding in one pump room from impacting the 1 adjacent pump room (s).

5.3 In the event of a CCW/SWC break in the CCW area of the SEB, if a pump room hatch is open, the flood would fail the associated pump. If the leak is not isolated in 30 minutes and a pump room door is impaired, the associated pump would fail.

The following event trees were developed:

CCW/SWC piping failure in CCW area, Modes 1-4 (Fig. 5.a).

l CCW/SWC piping failure in CCW area, Modes 4, 5, 6(level <23'), (Figure 5.b).

1 30 i ;

CCW/SWC piping failure in CCW area, Mode 6(level >23'), (Figure 5.c).

Calculations for Figure 5.a:

CCW/SWC line break in CCW area of SEB during Modes 1-4.

Initiating event, Node IE = 1.23E-7/ day Where, IE = CCW/SWC piping failure frequency in rooms 009 - 014 and room 022 - 026

=

pipe sections x pipe failure frequency per section Node A = BIF x AOT/365 = 7.7E-3 Node A = Prob. of hatch impaired in the operable CCW train pump room BIF = 0.4 CCW pump room hatch AOT = 7 days Node B = BIF x AOT/365 = 7.7E-3 Node B = Prob. of hatch impaired in the Swing CCW pump room l BIF = 0.4 CCW pump room hatch AOT = 7 days t

Node C = BIF x AOT/365 = 4.8E-2 Node C = Prob. of door impaired in the operable Train CCW pump l room BIF = 2.5 CCW pump room door AOT = 7 days l Node D = BIF x AOT/365 = 4.8E-2 l Node C = Prob. of door impaired in the Swing CCW pump room l BIF = 2.5 CCW pump room door AOT = 7 days Node E = SE-2 (see HRA work Sheet 6) i i

' Node F = (BIF x AOT/365)"2 = 2.3E-3 l l

' l l

31

, l

Where, Node F = Prob. of doors impaired to 1 of 2 operable CCW pump rooms BIF = 2.5 CCW pump room door AOT = 7 days Node G = PRA Run, CCDP given loss of CCW.

Calculations for Finure 5.b:

CCW/SWC line break in CCW area of SEB during Modes 4, 5, 6(level <23').

Initiating event, Node IE = 1.23E-7/ day Where, IE = CCW/SWC piping failure frequency in rooms 009 - 014 and room 022 - 026

=

pipe sections x pipe failure frequency per section Node A = BIF x AOT/365 = 1.15E-2 Node A = Prob. of hatch impaired in the operable CCW train pump room BIF = 0.6 CCW pump room Hatch AOT = 7 days Node B = BIF x AOT/365 = 1.15E-2 Node B = Prob. of hatch impaired in the Swing CCW pump room BIF = 0.6 CCW pump room hatch AOT = 7 days Node C = BIF x AOT/365 = 1.38E-1 ,

Node C = Prob. of door impaired in the operable Train CCW pump room BIF = 7.2 CCW pump room door AOT = 7 days Node D = BIF x AOT/365 = 1.38E-1 Node C = Prob. of door impaired in the Swing CCW pump room

)

BIF = 7.2 CCW pump room door {

AOT = 7 days 32

.m .- _ __ __ . _. _ _ . _ _ _ _ - . . _ _ . _ ._ .. _.

l I

l i

l 1

1 Node E = SE-2 (see HRA work Sheet 6) ,

I Node F = (BIF x AOT/365)**2 = 1.9E-2 l

Where, Node F = Prob. of doors impaired to 1 of 2 operable CCW l pump rooms BIF = 7.2 CCW pump room door AOT = 7 days i

Node G = IE-3 (see HRA work Sheet 3 )

! Node H = PRA Run,2 of 3 charging pumps fail to operate for 24 hrs.

Calculations for Figure 5.c:

' I CCW/SWC line break in CCW area of SEB, Mode 6(Level >23'). l l Initiating event , Node IE = 6.15E-8/ day l

Where, Node IE = Prob. Line break in the operating train of CCW/SWC.

=

l # sections x failure rate per section (# pipe sections were )'

! divided by 2 since one train is out of service in this Mode) l = 1.23E-7/2 = 6.15E-8 Node DR1 = BIF x AOT/365 = 1.15E-2

l

! l DR1 = Prob. of the operable CCW pump room hatch impaired BIF = 0.6 CCW pump room hatch AOT = 7 days Node DR2 = BIF x AOT/365 DR2 = Prob. Of operable CCW pump room door impaired BIF = 7.2 CCW pump room door ,

AOT = 7 days l Node OP = SE-2 (see HRA work sheet 6 )

j Node OPl = 1.5E-3 (see HRA work sheet 4)

Node SFP = 8E-4 Where, SFP = Mechanical failure to open/close eight critical valves to align l 33

(

l SFP to cool RCS and to cross tie CCW to the other unit.

l See Section 1.c above.

6. CCW Line Break in SEB Room 017 CCW piping system failure in the ECCS common area of SEB (Room 017) is evaluated.

Train A CCW piping failure causing flooding in Train B ECCS pump room through an

)

i open hatch is the bounding case. The event trees were simplified by considering the hatch to SEB room 015 open, and HPSI 18 was not credited.

Assumptions:

l 6.1 Train B CCW line break in room 017 would fail Train B ECCS but does not impact Train A ECCS equipment.

6.2 In addition to failing Train A ECCS, Train A CCW line break in room 17 with !

floor hatch to SEB room 002 open would also result in failure of Train B HPSI i due to water pouring directly on the pump. (LPSI and CS pumps are not located l l directly under the hatch and are not subject to water impingement.) However, ;

with the door to SEB room 002 closed, if the leak is not isolated in 30 minutes, all l

remaining ECCS B pumps (i.e., CS P013, and LPSI P016) would fail due to flooding.

l The following event trees were developed:

CCW line break in room 017, Modes 1-4, (Figure 6.a).

4 CCW line break in room 017, Modes 4, 5, 6(level <23'), (Figure 6.b).

CCW line break in room 017, Mode 6(level >23'), (Figure 6.c).

l l

Calculations for Figure 6.a:

CCW line break in room 017, Modes 1-4.

Initiating event, Node IE = 2.34E-8/ day Where, IE = CCW Train A piping failure frequency in room 017

! = # sections x failure rate per section

, 34 l

I l

Node HAT =BIF x AOT/365 = 8E-3 i HAT = Prob. Hatch impaired in room 002 BIF = 0.4 Hatch to room 002 I AOT = 7 days Node OP = 5E-2 (see HRA work sheet 7)

Node CDA = PRA Run, CCDP given loss of CCW A, HPSI 18, and ECCS B.

Node CDB = PRA Run, CCDP given loss of CCW A, and HPSI 18&l9.

Calculations for Figure 6.b: ,

1 CCW line break in room 017, Modes 4, 5,6(level <23'). The event tree was simplified by not crediting ECCS Train B pumps if the hatch to room 2 was impaired.

Initiating event, Node IE = 2.34E-8/ day 1

Where, IE = CCW Train A piping failure frequency in room 017

= # sections x failure rate per section Node HAT =BIF x AOT/365 = 1.15E-2 i

HAT = Prob. Hatch impaired in room 002 i BIF = 0.6 Hatch to room 002 AOT = 7 days )

Node OP = IE-3 (see HRA work sheet 3)

Node CHR = PRA Run,2 of 3 charging pumps fail to operate for 24 hrs.

Calculations for Figure 6.c:

CCW line break in room 017, Mode 6(level >23'). The event tree was simplified by not crediting ECCS B pumps if the hatch to room 2 was impaired.

Initiating event, Node IE = 2.34E-8/ day Where, IE = CCW Train A piping failure frequency in room 017

= # sections x failure rate per section R

35 i

l l

I l

Node HAT =BIF x AOT/365 = 1.15E-2 ,

i HAT = Prob. Hatch impaired in room 002 BIF = 0.6 Hatch to room 002 i AOT = 7 days l

Node OP = 1.5E-3 (see HRA work sheet 4) l Node SFP = 8E-4 I l

Where, SFP = Mechanical failure to open/close eight critical valves to align !

SFP to cool RCS and to cross tie CCW to the other unit.

See discussion for Figure 1.e above.

7. CCW Surge Tank Rupture CCW surge tank failure in both trains is evaluated in conjunction with impairment of their ;

associated barriers. To simplify the event trees, HPSI 18 was not credited in the calculations for Figures 7.a, 7.b, and 7.e, below.

Assumptions:

7.1 CCW Train B surge tank break would cause failure of Train A HPSI 17 pump if l floor plug in the SEB room 020 were open. The failure is due to water falling on the pump, but the other Train A ECCS pumps would not be impacted due to their locations relative to the hatch and flood draining to the SDC piping tunnel. If the floor plug in SEB room 020 is closed and the tank room door is open, water would flow to room 017 and down the stairs to elevation -15' 6" to the SDC tunnel. If the hatch from room 017 to room 015 is open, HPSI 18 pump would fait due to waterimpingement.

l 7.2 CCW Train A surge tank break would cause flooding in room 017 if the tank room door were open. If floor plug to room 002 were open, water would fall on HPSI 19 and fail the pump.

7.3 Although the tank rupture is considered selflimiting in the design basis of SONGS, it was included in this analysis. If the CCW Train A surge tank break is not

, isolated in 30 minutes, all Train B ECCS pumps in SEB room 002 would also fail.

If floor plug to room 015 were open, HPSI 18 pump would fail.

The following event trees were developed:

36 l

CCW B tank rupture, Modes 1-4, Figure 7.a.

CCW A tank mpture, Modes 1-4, Fig. 7.b.

CCW B tank rupture, Modes 4, 5, 6(level <23'), Fig. 7.c.

CCW A tank rupture, Modes 4,5,6(level <23'), Fig. 7.d.

CCW A tank rupture , Mode 6(level >23'), Fig. 7.e.

In Mode 6 (level >23'), the bounding case is CCW A tank rupture causing failure of Train B SDC. The CCW B surge tank rupture does not fail Train A SDC.

Calculations for Figure 7.a:

CCW Train B tank failure, Modes 1-4. The event tree was simplified by assuming the hatch to HPSI 18 pump room 015 to be impaired, and HPSI 18 was not credited.

Initiating event, Node IE = 1.2E-5/ day where, IE = Tank failure frequency (

Reference:

NUCLARR)

Node HAT = BIF x AOT/365 = 7.7E-3 HAT = Prob. of hatch to room 005 impaired.

BIF = 0.4 hatch from SEB room 020 to 005 l

AOT = 7 days Node CCW = PRA Run, CCDP given loss of CCW B and HPSI 17 & 18.

Calculations for Figure 7.b:

CCW Train A tank failure, Modes 1-4.

Initiating event, Node IE = 1.2E-5/ day where, IE = Tank failure frequency Node HAT = (BIF, x AOT/365)x(BIF2 x AOT/365) = 3.7E-4 HAT = Prob. of hatch to room 002 and door to Rm. 021 impaired.

BIF, = 0.4 hatch from SEB room 017 to 002 AOT = 7 days BIF 2= 2.5 door to Rm. 021 37

Node OP= SE-2 (see HRA work sheet 8)

Node CCB = PRA Run, CCDP given loss of CCW A and HPSI 18 and 19.

Node CCC = PRA Run, CCDP given loss of CCW A, ECCS B, and HPSI 18.

Calculations for Figure 7.c:

CCW Train B tank failure, Modes 4,5,6(level <23').

Initiating event, Node IE = 1.2E-5/ day where, IE = Tank failure frequency Node HAT = BIF x AOT/365 = 1.15E-2 HAT = Prob. Of hatch to room 005 impaired.

BIF = 0.6 hatch from SEB room 020 to 005 AOT = 7 days Node SDC = PRA Run, SDC A operates for 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months .

Node OP = IE-3 (see HRA work sheet 3)

Node CHG = PRA Run,2 of 3 charging pumps fail to operate for 24 hrs.

Node HPI = PRA Run, HPSI 18 operates for 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months .

Calculations for Ficure 7.d:

CCW Train A tank failure, Modes 4,5,6(level <23').

l l

Initiating event, Node IE = 1.2E-5/ day l where, IE = Tank failure frequency I Node DOR = BIF x AOT/365 = 1.4E-1 HAT = Prob. of door to room 021 impaired.

BIF = 7.2 door to SEB room 021 AOT = 7 days Node HAT = BIF x AOT/365 = 1.15E-2 l

HAT = Prob. of hatch to room 002 impaired.

38 I 1

I

BIF = 0.6 hatch from SEB roc n 017 to 002 4

AOT = 7 days Node OPl = SE-2 (see HRA work sheet 8)

Node OP2 = 1E-3 (see HRA work sheet 3)

Node CHR = PRA Run,2 of 3 charging pumps operate for 24 hrs.

Node HAl = BIF x AOT/365 = 1.15E-2 HAl = prob. of hatch to room 015 impaired BIF = 0.6 hatch to room 015 (HPSI 18 pump room)

AOT = 7 days Calculations for Figure 7.e:

CCW Train A tank failure, Mode 6(level >23').

Initiating event, Node IE = 1.2E-5/ day where, IE = Tank failure frequency Node DOR = (BIF xAOT/365)x(BIF 2xAOT/365) = 1.6E-3 i

HAT = Prob. of hatch to room 002 and door to Rm. 021 impaired.

BIF i= 0.6 hatch to SEB room 002 -

AOT = 7 days BIF 2= 7.2 door to Rm. 021 l

Node OPl = 5E-2 (see HRA work sheet 8)

Node SDC = not related to barrier impairment, not analyzed.

Node OP2 = 1.5E-3 (see HRA work sheet 4)

Node SFP = 8E-4 Where, SFP = Mechanical failure to open/close eight critical valves to align SFP to cool RCS and to cross tie CCW to the other unit.

See discussion for Figure 1.e above.

39 f

1

l l

8. SDC Line Break in Mode 4 A postulated SDC line break in the SDC tunnel, the SEB, or the Penetration Building while operating on SDC in Mode 4 with RCS temperature >212 F is evaluated. This operating condition occurs typically for about 3 hours3.472222e-5 days 8.333333e-4 hours 4.960317e-6 weeks 1.1415e-6 months in Mode 4 during SDC operation while cooling down to enter Mode 5. A line break would spill water which would l partially flash to steam creating a harsh environment. The flooding resulting from a SDC line break would accumulate in the SDC tunnel, Train B ECCS pump room, or SDC heat exchanger rooms and not impact other equipment. Since this mode of operation occurs immediately following power operation, barrier impairment frequencies ass Wated with power operation were used in calculating impairment probabilities.

The event trees were simplified and HPSI 18 was not credited.

Assumptions:

8.1 A SDC line break fails both trains of SDC regardless of barrier impairments, because the break may occur in common piping. -

8.2 Train A ECCS pump room is not protected by any barriers against steam release from a SDC line break in the SDC tunnel.

8.3 A SDC line break in Train B ECCS pump room does not impact Train A ECCS if the Train B pump room barriers are closed.

8.4 Design Basis of SONGS 2/3 credit operator action to isolate a SDC line break in 10 minutes while operating in Mode 4 with RCS temperature >212 F. Failure to isolate SDC line break in 10 minutes would lead to conditions outside the design basis and is assumed to lead to core damage regardless of barrier impairments.

Given that the SDC line break is isolated in 10 minutes, any equipment in the SEB exposed to the steam environment would fail.

8.5 Given that a SDC line break is isolated in 10 minutes, RCS water level will remain above the top of the hot leg and natural circulation cooling will be established using the steam generators. Eventually (assumed 5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months in this analysis) inventory makeup by HPSI pumps or charging will restore level in the pressurizer.

8.6 Steam release due to a SDC line break in the ECCS area of SEB propagating through doors 015 or 016, through the CCW pump room doors to the CCW pumps is considered a tortuous path.

8.7 It is conservatively assumed that one reactor trip per year results in a Mode 4 entry and SDC operation. Based on plant data, typical duration of SDC operation with 40

1 l

l RCS temperature >212 F while cooling down was estimated to be 3 hours3.472222e-5 days 8.333333e-4 hours 4.960317e-6 weeks 1.1415e-6 months . A conservative value of 6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months for cooldown was used in this analysis. This )

information was used in the SDC line break (steam release) event tree. l l

The following event trees were developed:

SDC line break in the SDC tunnel, ECCS A pump room, Penetration Building, or

, SDC heat exchanger rooms during SDC operation in Mode 4 with RCS l tempera.se >212 F, Figure 8.a.

SDC line break in the ECCS B pump room during SDC operation in Mode 4 with RCS temperature >212 F, Figure 8.b.

Calculations for Figure 8.a:

l l SDC line break in the SDC tunnel, ECCS A pump room, Penetration Building, or SDC heat exchanger rooms during SDC operation in Mode 4 with RCS temperature >212 F.

l Initiating event, Node IE = 2.7E-3/ day Where, IE = frequency of reactor trip to Mode 4

= one trip per year or 2.7E-3 per day was assumed from plant experience.

Node DR = 1.8E-7 DR = prob. of SDC line break in SDC tunnel, Penetration Building, SDC l heat exchanger rooms, or ECCS A pump room in a 6 hour6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months period

= # pipe sections x (failure rate /section-day) x (6/24)

= 6.87 E-7 x 6/24 = 1.8E-7 Node OP = 0.0 (Operator failure to isolate break in 10 minutes) !

= operator action to isolate the break was assumed in 10 minutes since failure ,

to do so was assumed to result in core damage regardless of barrier impairments (see assumption 8.4 above).

Node DRB = See FT attached to Figure 8.a .

DRB = Prob. of barriers impaired to ECCS B or CCW B pump rooms !

Node LX = PRA Run, AFW System operates for 24 hrs.

41

)

1

Node OP1 = lE-4 (see HRA work sheet 9)

Node DX = PRA Run, one of 3 charging pumps oper: .e for 24 hrs.

(Note: as a minimum, one charging pump is used to raise RCS level !

consistent with assumption 8.5 above)

Node HPI = PRA Run, HPSI 19 operates for 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months .

(Note: HPSI 19 is used to raise RCS level consistent with assumption 8.5 I

above)

Calculations for Figure 8A SDC line break in the ECCS B pump room during SDC operation in Mode 4 with RCS temperature >212 F.

Initiating event, Node IE = 2.7E-3/ day Where, IE = frequency of reactor trip to Mode 4

= one trip per year or 2.7E-3 per day was assumed from plant experience.

Node DR = 4.6E-8 DR = prob. of SDC line break in ECCS B pump room in a 6 hour6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months period

= # pipe sections x (failure rate /section-day) x (6/24)

= 1.84E-7 x 6/24 = 4.6E-8 Node OP = 0.0 (Operator failure to isolate break in 10 minutes)

= operator action to isolate the break is assumed in 10 minutes since failure j to do so was assumed to lead to core damage regardless of barrier impairments (see assumption 8.4 above).

Node DRB = BIF x NB x AOT/365 = SE-2 DRB = Prob. of barriers impaired to ECCS B pump room BIF = 1.3 average BIF for one door and one hatch AOT = 7 Days AOT for internal barrier was applied j Node LX = PRA Run, AFW System operates for 24 hrs.

Node OPl = IE-4 (see HRA work sheet 9)

Node DX = PRA Run, one of 3 charging pumps operate for 24 hrs.

42

l

9. High Energy Line Break Near the Roof of SEB l

This event only applies to Modes 1-4 when steam lines and feedwater lines are high energy. Steam from a line break near the roof of SEB could enter SEB through the floor plugs over the SDC heat exchangers in SEB rooms 016 and 018.

Assumotions:

9.1 HELB (MSLB or MFLB) near elevation 35' over the SEB would release steam to the SEB if the hatches over the SDC heat exchanger rooms were open. Based on engineering judgement, flooding from the MSLB or MFLB is expected to collect i in the heat exchanger room vault and not impact equipment in other rooms. Steam l release from this source propagating through the SDC heat exchanger rooms through SEB room 017 through the SDC tunnel to the Train A ECCS pump room isjudged to be tortuous path.

9.2 Steam release from this event to the CCW pump rooms through doors 015 or 016 )

and CCW pump room doors isjudged to be tortuous path. l 9.3 Steam release from this event to CCW pump rooms through the SDC tunnel is judged to be tortuous path.

Calculations for Figure 9.a:

An ET was developed for HELB on the roof of SEB in Modes 1-4, (see Figure 9.a). This event tree was truncated and simplified because multiple barrier failures with low l probabilities ofimpairment would have to fail to result in failure of safety equipment in the SEB. The event tree was simplified and door to room 20 was assumed to be open.

Initiating Event, Node IE = 6.91E-6/ day ,

i Where, IE = Frequency of HELB on the roof of SEB (# sections x impairment .

frequency per section) l Node A = (BIF xAOT i/(24x365)) x (BIF 2xAOT /365) i 2 x 2 = 5.25E-5

! Node A = Prob. of barriers impaired to rooms 016 or 018 admitting steam to the ECCS common area of SEB '

BIF, = 0.4 Plugs over SDC heat exchangers in room 016,018 AOT = 12 hrs steam barriers on the exterior of SEB BIF 2= 2.5 doors to rooms 016,018 AOT 2= 7 days interior doors A factor of two above represents barriers for two rooms 016/018.

43

Node B = (BIFxNBxAOT/365) = IE-1 Node B = Prob. ofimpairment of doors 015 or 016.

BIF = 2.5 Door 015 and 016 NB = 2 Doors 015 and 016 AOT = 7 days Node C = (BIFx AOT /365) = 7.7E-3 Node C = Prob. ofimpairment of hatch to CCW A pump room.

BIF = 0.4 hatch AOT = 7 days Node D = BIF x AOT/365 = 7.7E-3 Node D = Prob. ofimpairment of the hatch to CCW B pump room.

BIF = 0.4 hatch to CCW B pump room AOT = 7 days Node E =(BIF i xNB ixAOT/365) x (BIF 2xNB 2xAOT/365) = 2.5E-3 Node E = Prob. ofimpairment ofBarriers to ECCS pump rooms.

l BIF = 1.3 i

average BIF one door and one hatch to room 002.

NB = 2i one door and one hatch AOT = 7 days BIF2= 1.3 average BIF one door and one hatch to room 005. i NB2 = 2 one door and one hatch !

10. Fire Suppression Line Break in SEB Fire suppression line break in SEB rooms 022 to 026 may cause a failure of CCW pumps if the hatches above the pumps were impaired. Pump failure would be due to water falling on the pump. l Fire suppression line break in SEB room 017 could result in flooding of the ECCS B pump room if the hatch to room 002 were open.

Assumotions:

10.1 Flooding due to failure of the fire suppression line in SEB rooms 022 to 026 is detected and isolated with no impact on equipment except when the hatches over CCW pumps are open.

10.2 To simplify some of the event trees for the fire suppression line breaks in rooms 022 to 026, failure of two of three CCW pumps were conservatively treated as a total loss of CCW.

44 l

1

1 l

10.3 Flooding due to failure of the fire suppression line in room 017 would result in failure of HPSI 19 pump if the hatch to room 002 were open. Failure would be caused by water falling on the pump. If flooding is not isolated in one hour, the :

entire Train B ECCS is conservatively assumed to fail. HPSI 18 pump is not I credited in this case.

The following event trees were developed:

)

Fire suppression line break in rooms 022 to 026, Modes 1-4, Figure 10.a. !

Fire suppression line break in room 017, Modes 1-4, Figure 10.b.

Fire suppression line break in rooms 022 to 026, Modes 4, 5, 6(level <23'), Figure ;

10.c. I i

1 Fire suppression line break in rooms 022 to 026, Mode 6(level >23'), Figure 10.d.

l 1

Fire suppression line break in room 017, Modes 4, 5, 6(level <23'), Figure 10.e.

Fire suppression line break in room 017, Mode 6(level >23'), Figure 10.f.

Calculations for Figure 10.a Fire suppression line break in rooms 022 to 026, Modes 1-4.

Initiating event, Node IE = 1.2E-6/ day 1 where, IE = Frequency of fire suppression line break in rooms 022 to 02. l

(# sections x pipe failure frequency per section) !

I Node A = BIF x AOT/365 = 7.8E-3 Node A = prob. of hatch impaired to CCW A pump room.

BIF = 0.4 CCW A pump room hatch AOT = 7 days )

Node B = BIF x AOT/365 = 7.8E-3 Node B = prob. of hatch impaired to CCW B pump room.

BIF = 0.4 CCW B pump room hatch AOT = 7 days Node C = BIF x AOT/365 = 7.8E-3 45 l

r l

Node C = prob. of hatch impaired to Swing CCW pump room.

BIF = 0.4 Swing CCW pump room hatch AOT = 7 days Node D = PRA Run, CCDP given loss of 2 of 3 CCW pumps.

Node F = PRA Run, CCDP given loss of CCW.

Calculations for Figure 10.b:

Fire suppression line break in room 017, Modes 1-4.

Initiating event, Node IE = 5.6E-7/ day where, IE = fire suppression piping failure rate in room 017. (# sections x failure rate per section)

Node A = BIF x AOT/365 = 7.7E-3 l BIF = 0.4 Prob. Of hatch impaired in room 002 AOT = 7 days Node B = IE-2 (see HRA work sheet 10)

Node C = PRA Run, CCDP given loss of ECCS B. I l

Calculations for Figure 10.c:

Fire suppression line break in rooms 022 to 026, Modes 4, 5, 6(level <23').

Initiating event, Node IE = 1.2E-6/ day where, IE = Frequency of fire suppression line break in rooms 022 to 026. (#

sections x pipe failure frequency per section)

Node A = BIF x AOT/365 = 1.15E-2 Node A = prob. of hatch impaired to CCW A pump room.

BIF = 0.6 CCW A pump room hatch AOT = 7 days Node B = BIF x AOT/365 = 1.15E-2

' Node B = prob. of hatch impaired to CCW B pump room.

BIF = 0.6 CCW B pump room hatch AOT = 7 days i

Node C = BIF x AOT/365 = 1.15E-2 i

46 i

l

! Node C = prob. of hatch impaired to Swing CCW pump room.

BIF = 0.6 Swing CCW pump room hatch l

AOT = 7 days Node D = IE-3 (see HRA work sheet 3) l Node E = PRA Run,2 of 3 charging pumps fail to operate for 24 hrs.

l Calculations for Figure 10.1 l

Fire suppression line break in rooms 022 to 026, Mode 6(level >23').

Initiating event, NMe IE = 1.2E-6/ day where, IE = Frequency of fire suppression line break in rooms 022 to 026. (#

l sections x pipe failure frequency per section)

Node HAT = BIF x AOT/365 = 1.15E-2 Node A = prob. of hatch impaired to Operable CCW pump room.

BIF = 0.6 operable CCW pump room hatch l AOT = 7 days Node OP = 1.5E-3 (see HRA work sheet 4)

Node SFP = 8E-4 Where, SFP = Mechanical failure to open/close eight critical valves to align

, SFP to cool RCS and to cross tie CCW to the other unit.

I See Section 1.c above.

Calculations for Figure 10.e:

l Fire suppression line break in room 017, Modes 4, 5, 6(level <23').

i Initiating event, Node IE = 5.6E-7/ day where, IE = fire suppression piping failure rate in room 017. (# sections x failure rate per section) i- Node A = BIF x AOT/365 = 1.15E-2 BIF = 0.6 Prob. of hatch impaired to room 002 AOT = 7 days l Node B = IE-2 (see HRA work sheet 10) 47 I

l I

i

Node C = PRA Run, SDC A operates for 24 hrs.

Node D = IE-3 (see HRA work sheet 3)

Calculations for Figure 10.f:

Fire suppression line break in room 017, Mode 6(level >23').

Initiating event, Node IE = 5.6E-7/ day where, IE = fire suppression piping failure rate in room 017. (# sections x failure rate per section)

Node A = BIF x AOT/365 = 1.15E-2 BIF = 0.6 Prob. ofhatch impaired to room 002 AOT = 7 days Node B = IE-2 (see HRA work sheet 10)

Node C = 1.5E-3 (see HRA work sheet 4)

Node SFP = 8E-4 Where, SFP = Mechanical failure to open/close eight critical valves to align SFP to cool RCS and to cross tie CCW to the other unit. .

I See discussion for Figure 1.e above.

48

BARRIER IMPAIRMENT RISK

SUMMARY

TABLES SAFETY EQUIPMENT BUILDING inmator ITEMS INITIATOR (Modes 1-4)

8'"*0*

Frequency Per File Name Day Freqm Per Day 1 Fig 1.a CWS Flood without Floodwatch, Modes 1-4 3.4E-05 1.53E-11 SO2CW1 2 Fig 1.b CWS Flood with Floodwatch, Modes 1-4 3.40E-05 1.29E-10 SO2CWS 3 Fig 2.a AFW Turbine Steam Une Break, Modes 1-4 2.14E-07 8.61E-15 SO2AF2 4 Fig 3.a Auxiliary Steam Line Break in Corridor, Modes 1-4 2.14E-07 3.44E-13 SO2SL1 5 Fig 4.a High Energy Line Break in Turbine Building, Modes 1-4 1.90E-05 1.12E-11 SO2HE1 6 Fig 5.a CCW/SWC Line Break, CCW Area of SEB, Modes 1-4 1.23E-07 5.15E-14 SO2CC1 7 Fig 6.a CCW Line Break in Room 17 of SEB, Modes 1-4 2.34E-08 6.33E-15 SO2CC2 i 8 Fig 7.a CCW Gurge Tank Rupture Train B, Modes 1-4 1.2E-05 5.90E-13 SO2CC3 !

l 9 Fig 7.b CCW Surge Tank Rupture Train A. Modes 1-4 1.2E-05 3.82E-14 SO2CC8 !

10 Fig 8.a Reactor Trip / Steam Release in SEB: SDC Line Break during Mode 4 2.70E-03 1.07E-13 SDRSB4 11 Fig 8.b Reactor Trip / Train B SDC Une Break in ECCS B Pump Room, Mode 4 2.70E-03 2.45E-15 SO2SD1 !

12 Fig 9.a MSLB/FWLB on 35' elevation in SEB, Modes 1-4 4.90E-06 1.37E-12 SO2SL7 l

13 Fig 10.a Fire Suppression Line Break in Room 22-26, Modes 1-4 1.20E-06 1.26E-15 I SO2FR4 i

14 Fig 10.b Fire Suppression Line Break in Room 17, Modes 1-4 5.6E-07 6.04E-17 SO2FR7 MODES 1 - 4 CDF 1.58E-10 nsg\ barrier \sebcdr.wb2 '

i l

i f ITEMS INITIATOR (Modes 4,5 and 6 level <23*) Initiah Core Darnage Flie Name l Frequency Frequency l

15 Fig 1.c CWS Break with Floodwatch, Modes 4,5 and 5(level <23') 2.4E-05 1.63E-10 SO2CW6 l

i 16 Fig 1.d CWS Break without Floodwatch, Modes 4, 5 and 6(level <23') 2.40E-05 1.91E-11 SO2CW3 17 Fig 3.b Auxiliary Steam Line Break in Corridor, Modes 4, 5 and 6(level <23') 2.14E-07 2.91E-12 SO2SL3 18 Fig 4 b High Energy Line Break in Turbine Building, Modes 4, 5 and 6(level <23') 3.2E-06 2.81E-11 SO2SL5 j 19 Fig 5.b CCW/SWC Line Break in CCW Area, Modes 4,5 and 6(level <23') 1.23E-07 2.73E-13 SO2CC9 20 Fig 6.b CCW A Line Brook in Room 17, Modes 4,5 and 6(level <23') 2.34E-08 5.22E-13 SO2CC7 21 Fig 7.c CCW B Tank Rupture, Modes 4,5 and 6(level <23*) 1.20E-05 8.12E-13 SDRCCS 22 Fig 7.d CCW A Tank Rupture , Modes 4,5 and 6(level <23*) 1.2E-05 9.76E-13 SO2CCS i 23 Fig 10.c Fire Suppression Line Break in Rooms 22-26, Modes 4, 5 and 6(level <23') 1.2E-06 9.16E-13 SO2FR2 24 Fig 10.e Fire Suppression Line Break in Room 17, Mode 4,5 and 6(level <23') 5.60E-07 1.98E-16 SO2FRS ,

MODES 4,5 AND 6 (LEVEL <23*) 2.17E-10 nsg\ barrier \sebedr.wb2

i I

Initiator Core ramage l ITEMS INITIATOR (Modes 6 level >2T) ; File Name Frequency Frequency g 25 Fig 1.e CWS Break with Floodwatch, Modes 6(level >23*) 2.40E-05 1.93E-10 SO2CW7 l

l 26 Fig 1.f CWS Break without Floodwatch, Mode 6(leve!>23') 2.40E-05 2.27E-11 SO2CW4 l

27 Fig 3.c Auxiliary Steam Une Break in Corridor, Mode 6(level >23') 2.14E-07 7.60E-12 SO2SL4 28 Fig 4.c Steam Line Break in 30' Turbine Building Mode 6(level >23') 3.20E-06 8.18E-11 SO2SLS 29 Fig 5.c CCW/SWC Line Break in CCW Area, Modes 6(level >23') 6.34E-08 2.59E-12 SO2C10 30 Fig 6.c CCW Line Break in Room 17 Modes 6(level >23') 2.34E-08 6.19E-13 SO2CC6 s 31 Fig 7.e CCW A Suego Tank Rupture, Modes 6(level >23') 1.20ti45 2.21E-12 SO2CC4 32 Fig 10.d Fire Suppression Line Break in Rooms 22-26, Mode 6(leve!>23') 1.20E-06 3.17E-11 SO2FR3 33 Fig 10.f Fire Suppression Line Break in Rooms 17, Mode 6(level >23') 5.60E-07 1.48E-13 SO2FRS MODES 6 LEVEL >23' 3.42E-10 ;

nsg\ barrier \sebcdr.wb2

ab m. s -Jm s-ss--- J x -*n.- e -a no a u--- --1-a s

.+ - u-. -- -- - - - - -- -

't APPENDIX B PlPE FAILURE CALCULATIONS SAFETY EQUIPMENT BUILDING t

PIPE FAILURE CALCULATIONS ,

For SEB Pipe sections were counted based on plant walk-downs and review of P& ids. Pipe section definitions is consistant with the definition in EPRI report: EPRI TR-102266, " Pipe Failure Study Update", dated April 1993.

Non-safety class piping is assumed to have a failure rate ten times that of safety class piping of similar size. This is shown as Safety Factor (EF) of 10 in this calculation.

( CCW AND SWC Table 1 Safety Equipment CCW/SWC Pipe Failure Area Room Hazard Drawing / Level Sections Failure rate

Per Day 2-SE-(-15)-136 17 6" 83016 7 2.34E-8 2-SE-(-5)-135A 22-26 2HX, 6" 83016 33 1.10E-7 2-S5-(-5)-135A 9-14 83015 4 1.33E-8 Failure Rate per Day =(1.39E-10 Failures /Section-Hr.)(24Hr./ Day)(# of Sections)

CCW/SWC LB Calculation SEB ROOM 17:

1.39E-10/Hr.X24Hr./DayX7 Sections = 2.34E-8/ day SEB ROOMS 9-14 and 22-26:

1.39E-10/Hr.X24Hr./DayX33 Sections =1.10E-7/ day 1.39E-10/Hr.X24Hr./DayX4 Sections =1.33E-8/ day

Total 1.10E-7 + 1.33E-8 = 1.23E-7/ day TURBINE DRIVEN AUX FEED. STEAM SUPPLY LINE..

TABLE 2 Auxiliary Feedwater Pump Room Area Room Hazard Drawing Sections Failure Rate l

Per Day 2-CT-(-2)-142B 6" AFLB 83016 10 2.14E-7 Failure Rate Per Day =(8.9E-10 Failure / Sections-Hr.)(24 Hr/ Day)(# of Sections) i Auxiliary Feedwater Pump Room LB Calculation.

8.9E-10/HrX74Hr./DayX10 Sections = 2.14E-7/ day l

2 l

l l

_- . . - - = - .- . _ - _ -

4 l

ASLB IN THE CORRIDOR

> Table 3 Auxiliary Steam Line Break in Corridor Next to Unit 2 SEB Area Room Hazard Drawing / Level Sections Failure Rate

Per Day 2-P E-9-2A 111 83001 2-PE-(-18)-2B 110 10" 83002 2.14E-7 2-PE-30-2C, D 113 ASL 83028 1 2-PE-30-2D 207 208 83029 3-PE-9-2A 209-214 40169C 3-PE-(-18)-28 40169D 3-PE-30-2C,D Failure Rate Per Day =(8.9E-10 Failure / Sections-Hr.)(24 Hr/ Day)(# of Sections)

Safety Factor (SF) = 10 Auxiliary Steam LB Calculation 8.9E-10/Hr.X24Hr./DayX10X1 Sections =2.14E-7/ day FIRE SUPPRESSION llNE BREAK IN THE SEB Table 4 ROOMS 17, AND 22-26 Area Room Hazard Drawing / LEVEL Sections Failure Rate !

Per Day 2-SE-(-5)-135A 22-26 4" FSLB S O23-403-24-319 12 1.122E-6 2-SE-(-5)-135A 22-26 1"FSLB S O23-403-24-319 76 1.1 E-6 2-SE-(-15)-136 17 4" FSLB S O23-403-24-335 10 4.3E-7 2-SE-(-15)-136 17 1"FSLB S O23-403-24-335 32 4.62E-7 Failure Rate Per Day =(3.98E-10,1.44E-8 Failure / Sections-Hr.)(24 Hr/ Day)(# of Sections) Respectively.

Fire Suppression LB Calculation.

3.98E-10/Hr.X24Hr./DayX12 Sections =1.15E-7/ day 6.01E-10/Hr.X24Hr./DayX76 SECTIONS = 1.1E-6/ day TOTAL Rooms 22-26 1.15E-7 + 1.1E-6 = 1.22E-6/ day 3.98E-10Hr.X24Hr./DayX10 Sections = 9.55E-8/ day 6.01E-10Hr.X24Hr./DayX32 (Heads) = 4.62E-7/ day TOTAL Room 17: 9.55E-8 + 4.62E-7 = 5.58E-7/ day 3

SDC Line Break in ECCS A Pump Room. SDC Tunnel. SDC Heat Exchanaer and Penetration Buildina TABLE 5 l

Area Room Hazard Drawing Sections Failure Rate

Per Day 2-SE-(-15)-136 > 6" SDCLB 83015 149 6.87E-7 l 401128 Failure Rate Per Day =(1.92E-10 Failure / Sections-Hr.)(24 Hr/ Day)(# of Sections)

SHUTDOWN COOLING in Safety Equipment Room LB Calculation 1.92E-10/Hr.X24Hr./DayX149 Sections = 6.87E-7/ day SDC Line in ECCS B Pump Room (SEB Room 002)

Table 6 Area Room Hazard Drawing / Level Sections Failure Rate ;

Per day 2-SE-(-15)-137A 002 12"SDCLB 40112B 40 1.84E-7 83015 Failure Rate Per Day =(1.92E-10 Failure / Sections-Hr.)(24 Hr/ Day)(# of Sections) ;

1.92E-10/Hr.X24Hr./DayX40 Sections = 1.84E-7/ day I l

l 1

I 4

HELB in Turbine Buildina ASLB in The Turbine Buildina Table 7 Auxiliary Steam Line Break in Turbine Building Area Room Hazard Drawing / Level Sections Failure rate

Per Day 2-PE-9-2A 111 83001 2-PE-(-18)-2B 110 83002 2-PE-30-2C,D 113 10" 83028 15 3.20E-6 3-PE-9-2A 207-209 ASL 83029 3-PE-(-18)-2B 211-214 40169C 3-PE-30-2C,D 40169D Failure Rate per Day =(8.9E-10 Failures /Section-Hr.)(24Hr./ Day)(# of Sections)X10.

Safety Factor (SF) =10 8.9E-10/HrX24Hr/DayX10X15 Section = 3.20E-6/ day ;

l 1

BDLB in The Turbine Buildina.

TABLE 8 Blowdown Line Break Area Room # Hazard Drawing / Level Sections Failure Rate Per Day 2-PE-9-2A 111 83001 2-PE-(-18)-2B 110 83002 2-PE-30-2C, D 113 8" ASL 83028 17 2.3E-6 !

2-PE-30-2D 207,208 83029 ;

3-PE-9-2A 209-214 40141B 3-PE-(-18)-2B ;

3-PE-30-2C,D l Failure Rate Per Day =(5.64E-10 Failure / Sections-Hr.)(24 Hr/ Day)(# of Sections)

SAFETY FACTOR (SF)=10 5.64E-10/Hr.X24Hr./DayX10X17 Sections = 2.3E-6/ day l

5 l l

t 1 MFWLB in the Turbine Buildina TABLE 9 j Main Feed Water Line Break i Area Room # Hazard Drawing Sections Failure Rate Per Day 2-P E-9-2A 111 MFWLB 83002 1 1.54E-7 ;

2-PE-(-18)-2B 110 20" 83029 l 2-PE-30-2C, D 113 40156B 3-PE-9-2A 207-9 40156B-S03 3-PE-(-18)-2B 211-14 2-PE-30-20, D Failure Rate Per Day =(6.40E-10 Failure / Sections-Hr.)(24 Hr/ Day)(# of Sections) !

SAFETY FACTOR (SF) =10 l

6.4E-10/Hr.X24Hr./DayX10X1 Section = 1.54E-7/ day l l MSLB in the Turbine Buildina. 1 TABLE 10 i Main Steam Line Break Area Room # Hazard Drawing Sections Failure Rate Per Day l

2-P E-9-2A 111 > 10' 83001 61 1.31 E-5 2-PE-(-18)-2B 110 MSLB 83002 2-PE-30-2C, D 113 83028 3-PE-9-2A 207-9 83029 3-PE-(-18)-2B 211-14 40141E 3-PE-30-2C, D 40141F 1

l Failure Rate Per Day =(8.9E-10, Failure / Sections-Hr.)(24 Hr/ Day)(# of Sections) X10 j SAFETY FACTOR (SF) =10 8.9E-10/Hr.X24Hr./DayX10X61 Sections =1.31E-5/ day TOTAL HELB in Turbine Building:

3.2E-6 + 2.3E-6 + 1.54E-7 + 1.31E-5 = 1.9E-5/ day 9

6 l

.- - -- _ . . . -- . _. -=

HELB On the Roof of SEB MSLB On the Roof of SEB Table 11 MAIN STEAM LINE BREAK l Area Room Hazard Drawing / Level Sections Failure Rate j # Per Day 2-PE-45-3A 306 40" MSL 83002,83029 50 1.07E-6 2-PE-63-3B 406 40123A,D,F 40123A,D,FSO3 Failure Rate Per Day =(8.9E-10 Failure / Sections-Hr.)(24 Hr/ Day)(# of Sections)

Main Steam LB Calculation l 8.9E-10/Hr.X24Hr./DayX50 Sections =1.07E-6/ day l

MFWLB On the Roof of SEB Table 12

MAIN FEEDWATER LINE BREAK t ,

l Area Room Hazard Drawing / Level Sections Failure Rate l

Per Day l )

! 2-PE-45-3A 306 20" 83002,83029 38 5.84E-6 2-PE-63-38 406 MFWLB 401568, 40156BSO3 l Failure Rate Per Day =(6.4E-10 Failure / Sections-Hr.)(24 Hr/ Day)(# of Sections) i Safety Factor (SF) =10 Main Feedwater LB Calculation l

6.4E-10/Hr.X24Hr./DayX10X38 Sections =5.84E-6/ day HELB On the Roof of SEB MSLB + MFWLB = 1.07E-6 + 5.84E-6 = 6.91E-6/ day L

l 7

1

f f

1 APPENDIX C HRA WORK SHEETS I

_ _ ._

ML20113C508

Text

References:

References:

References:

Reference:

Reference:

Reference:

SUMMARY

Navigation menu

Search